Relación entre satisfacción del cliente y calidad de atención

5.1 Introduction

5.1.1 Lung cancer

Globally, lung cancer was the most common cancer in 1985, with an estimated 896.000 new cases, or 11.8% of the total, about 61% of which occurred in developed countries (Parkin et al. 1993). However, there are wide geographical variations in risk

(Coleman et a l 1993) with estimated male age-standardised incidence rates ranging

from 2.5 per 100,000 in Western Africa to 73.6 in North America (Parkin et al. 1993).

In females, incidence rates are much lower, the highest estimated rate o f 28.9 per 100.000 also occurring in North America; worldwide, lung cancer is only the fifth most frequent cancer of women accounting for 5.8% of the total.

In European Community countries, the highest male incidence rates are demonstrated in Belgium, the Netherlands and the United Kingdom with the lowest rates in Portugal, Spain and France (Jensen et a l 1990). Amongst females, the lowest rates also occur

in the latter countries, but the highest incidence and mortality rates are apparent in the United Kingdom with high rates also in Ireland and Denmark. Within the United Kingdom, during the period 1985-88, the highest mortality rates for both genders occurred in Scotland, at 76.4 per 100,000 for males and 27.0 for females (La Vecchia

et al. 1991b). The lowest rates were demonstrated in Northern Ireland, with those in

100,000 respectively. In England, lung cancer is the commonest cancer in males, accounting for 32% of the total, whilst it ranks third in females at 10% (Roman & Carpenter 1995b).

Lung cancer was a relatively rare disease in the early years o f this century (Clemmesen 1965). Temporal trends since, and variations in geographic distribution, can be accounted for almost entirely by national cigarette smoking habits (Tomatis et

al. 1990). However, a number of occupational associations have also been suggested

(Clemmesen 1965. Roman & Carpenter 1995b). In the sixteenth century, von Hohenheim and Georg Bauer noted a peculiar pulmonary disease, termed ’mala metallorum’, amongst metal miners in the Erz mountain region of Eastern Europe (Clemmesen 1965). This was identified as a malignant neoplasm in 1879. During this century, increased frequency of lung cancer in miners has been reported, especially in those mining uranium and other metals (Tomatis et al. 1990). This has been associated

with exposure to radon and its decay products (Moolgavkar et a/. 1993). High levels

of radon exposure in residential homes have also been implicated with increased risk for lung cancer amongst females in the Austrian Alps (Ennemoser et al. 1994).

Similarly, occupational exposure to asbestos results in an increased risk for lung cancer (Clemmesen 1965. Tomatis et al. 1990. Coggon et al. 1995). It has been estimated that

5.7% of all lung cancers in males in the west of Scotland were asbestos related (De Vos Irvine et al. 1993). Increased risk for lung cancer has also been noted amongst

occupations working with metals, such as metal polishers, fettlers and dressers, moulders and coremakers, metal drawers, and electroplaters who are exposed to

chromâtes (Tomatis et al. 1990. Roman & Carpenter 1995b. Coggon et a l 1995).

Although a high risk has been reported for labourers in coke ovens, who have a high exposure to poly cyclic aromatic hydrocarbons (Coggon et a l 1995), Sorahan et a l

(1994) reported only limited evidence of an occupational role in excesses of lung cancer amongst steel foundry workers in the United Kingdom. Similarly, although Leon et a l (1994) demonstrated an increased risk among newspaper printers exposed

to ink mist, the magnitude of the increase could be explained by confounding with smoking. Indeed, comment has been made that the strong confounding effects of smoking could affect the interpretation of increased risks for many occupations (Coggon et al 1995. Roman & Carpenter 1995b).

Air pollution has been considered to be a possible aetiological factor (Clemmesen 1965. OPCS 1994b), the risk for lung cancer being higher in urban than in rural areas (Tomatis et a l 1990). In the United Kingdom, Doll and Bradford Hill (1964a)

reported lower mortality rates among medical practitioners living in rural areas than in large towns, which could not be ascribed to differences in smoking habits. However, Lyon et a l (1980b) in a study comparing rural and urban Mormons in Utah,

most of whom were non-smokers, found no difference in lung cancer risk. Tomatis et

a l (1990) consider it probable that air pollution may contribute to lung cancer

mortality in heavily polluted areas. However, no evidence of increased risk was found in a study of residents living near ten industrial incinerators of waste solvents and oils in Great Britain (Elliott et a l 1992). Moreover, Doll and Bradford Hill (1964b) in

their study of medical practitioners concluded that the very low death rates from lung cancer in non-smokers of both genders living mainly in urban areas, did not suggest

that air pollution had been an important factor. It is of interest to note, therefore, that at the inception of a large case-control study in 1948 to investigate the rapidly increasing mortality from lung cancer in the United Kingdom, Doll considered that increases in cars and the tarring of roads were more likely factors than cigarette smoking (Peto 1994). However, smoking proved to be the only major difference between cases and controls.

The causal relationship between smoking, especially cigarette smoking, and lung cancer has subsequently been confirmed conclusively in both genders (Doll & Bradford Hill 1964a, 1964b. Doll et al. 1980. Doll et al. 1994). In England and

Wales, the proportion of lung cancer deaths attributable to smoking has been estimated at 94% and 80% for males and females respectively (lARC 1986). Time trends in incidence and mortality from lung cancer reflect the prevalence of cigarette smoking in different generations (Tomatis et al. 1990), with an estimated induction period of

at least twenty years (Clemmesen 1965. OPCS 1994b). Moreover, variations in lung cancer trends not only reflect changes in tobacco consumption but also alterations in the type of tobacco and tar content of cigarettes, and the use of filters (Wynder & Hoffman 1994). It is therefore considered that data for lung cancer incidence and mortality will provide acceptable and effective surrogate markers for variations in cigarette usage.

5.1.2 Liver cirrhosis

standardised mortality rates in the mid 1980s ranging from 1.4 per 100,000 in Iceland to 56.4 in the Korean Republic (WHO 1988). Female rates are generally lower than male, the lowest of 0.7 per 100,000 also occurring in Iceland; the highest at 17.9 per

100,000 is reported in Romania. Among European countries, high rates are recorded for both genders in Hungary, Portugal and Italy at 50.3, 33.4 and 32.1 per 100,000 for males and 17.4, 10.9 and 11.6 per 100,000 for females. The lowest male mortality rates, at 2.7 and 3.2 per 100,000 occur in the Irish Republic and Northern Ireland. Amongst females, the lowest rates are reported in the Irish Republic and Norway at 2.1 and 2.2 per 100,000, with a low rate of 2.7 in Northern Ireland, Sweden and Holland. Within the United Kingdom the highest mortality rates occur in Scotland, at 7.3 and 4.5 per 100,000 for males and females respectively. Rates of 4.5 per 100,000 for males and 3.1 for females in England and Wales are intermediate between those in Scotland and Northern Ireland.

There are many causes of liver cirrhosis (Richmond & Finlay son 1987), among the commonest being alcoholism, chronic active hepatitis due to hepatitis B, Delta and non-A non-B viruses, primary biliary cirrhosis and schistosomiasis (Eddleston 1990). Rare causes include Wilson’s disease and haemochromatosis. In the Middle and Far East, and many areas of Africa, chronic hepatitis B virus infection and schistosomiasis are the most important factors. However, chronic alcoholic liver disease is the commonest cause of cirrhosis in Europe (Saunders et al. 1981. Eddleston 1990), and

variations in mortality rates between countries are closely associated with differences in per capita levels of alcohol consumption (Smith 1981. Farrell & Strang 1992). In

primary biliary cirrhosis (Richmond & Finlay son 1987), 80% of cases admitted to district general hospitals are associated with alcohol abuse (Eddleston 1990).

A dose-response relationship between alcohol consumption and the risk of developing liver disease and cirrhosis has been demonstrated (Johnson & Williams 1985. Norton

et al. 1987). Consistent high daily alcohol consumption appears to be more likely to

cause cirrhosis than episodic ’binge’ drinking (Smith 1981. Richmond & Finlay son 1987), with a dose-duration period for the development of cirrhosis from five to over fifteen years (Royal College of Psychiatrists 1986. I ARC 1988). However, liver cirrhosis mortality has been shown to be extremely sensitive to changes in alcohol consumption, with rates rising or falling within one or two years of alterations in per

capita consumption (Smith 1981. Kendell 1984).

Liver cirrhosis mortality has been used as a marker of alcohol abuse in studies on alcoholism (Royal College of Psychiatrists 1979. Breeze 1985. Prior 1988. Paton 1994). Walsh and Walsh (1973) examining the validity of five indices of alcoholism concluded that there was a high and consistent correlation between per capita

measures of alcohol consumption and the death rate from cirrhosis. It was therefore considered appropriate to use liver cirrhosis mortality as a surrogate marker for alcohol consumption in this study. However, the sensitivity of cirrhosis mortality to changes in consumption when compared to the induction period for cigarette smoking and lung cancer o f at least twenty years must be borne in mind.

5.1.3 Aims

The purpose of this part of the study was to examine the temporal and spatial epidemiology of lung cancer mortality and incidence and liver cirrhosis mortality in England and Wales during this century.

5.2 Materials and Methods

Data for lung cancer mortality and incidence were obtained from the sources described in Chapter 3 and processed in the same manner. However, SIRs for geographic distribution were produced only for the period 1979-83.

No data were available for liver cirrhosis incidence. Mortality data, commencing in 1911-15, were obtained from the same sources and processed in the same manner as Chapter 3 for temporal trends. However, age-specific mortality rates are presented graphically by period of death. To assess geographic distribution, the number of liver cirrhosis deaths in each RHA were aggregated for the period 1974-78 (OPCS 1976- 80). National age-specific mortality rates were prepared for this period and Standardised Mortality Ratios (SMR) calculated, using the method o f indirect standardization described in Chapter 3.

5.3 Results

5.3.1 Lung cancer

5.3.1.1 Temporal trends

In the early part of the century, lung cancer mortality was low in both genders, with fluctuating rates until 1916-20 (Fig. 5.1). In that period the age-standardised rates stood at 1.63 per 100,000 for males and 0.86 for females (Table 5.1). Subsequently, there has been a profound and significant increase in both genders. This was more pronounced in males and continued until 1971-75, when the mortality rate peaked at 74.80 per 100,000. Since then there has been a steady decrease in the rate to 59.60 per

100,000 in 1986-90 and 55.00 in 1991. Similarly, male age-standardised incidence rates increased from 66.20 per 100,000 in 1962-66 to 77.60 in 1972-76, with a subsequent decrease to 67.00 in 1982-86 and 61.60 per 100,000 in 1989 (Table 5.2). In females, however, the mortality rate has continued increasing to 20.35 per 100,000 in 1986-90, although there has been a slight reduction to 19.65 in 1991 (Table 5.1). Moreover, female age-standardised incidence rates have steadily increased from 9.40 per 100,000 in 1962-66 to 20.20 in 1982-86 and 22.70 in 1989 (Table 5.2).

In both genders the highest rates for lung cancer incidence (Fig. 5.2, Table 5.3) and mortality (Fig. 5.3, Tables 5.4 & 5.5) occur in the older age groups. However, peak incidence and mortality rates occur in the 80-84 age group in males, and in the 75-79 group in females.

Males

Truncated age-standardised rates (Fig. 5.4) reveal increases in lung cancer mortality in the early part of the century which became far more marked after 1916-20. In males aged 35-64 years this continued until 1961-65, reaching a rate of 113.00 per 100,000 (Table 5.1). Subsequently, the mortality rate has progressively fallen to 67.80 per

100.000 in 1986-90, with a further reduction to 61.00 in 1991. Similarly, incidence rates have decreased from 119.00 per 100,000 in 1962-66 to 85.60 in 1982-86 (Fig. 5.5) and 73.60 in 1989 (Table 5.2).

In older males, however, mortality increased until 1976-80 reaching a peak rate of 624.00 per 100,000, before decreasing to 549.00 in 1986-90 (Fig 5.4) and 515.00 in 1991 (Table 5.1). Incidence rates in this group also showed increases between 1962-66 and 1977-81, from 411.00 to 612.00 per 100,000, before decreasing to 576.00 in 1982- 86 (Fig. 5.6) and 553.00 in 1989 (Table 5.2).

Both age-specific mortality rates (Fig. 5.7, Table 5.4) and incidence rates (Fig. 5.8, Table 5.3) confirm a decreasing trend in most age groups during the latter part o f this century. Only in those aged 85 or over had lung cancer mortality not started to decline by 1986-90, although the rate has decreased from 702.6 per 100,000 in that period to 659.1 in 1991 (Table 5.4). In males aged 35-39 mortality rates peaked in 1951-55, to be followed by peak rates in 1956-60 for those groups aged from 40-44 to 55-59. The subsequent decreases in these age groups could represent a partial period effect. The highest mortality rates for those aged 60-64 and 65-69 did not occur until 1966- 70, to be followed sequentially in 1971-75 for those aged 70-74, 1976-80 for those

aged 75-79 and in 1981-85 for those aged 80-84. Age-specific incidence rates have been decreasing since 1962-66 for all age groups from 35-39 to 55-59 (Table 5.3). However, peak rates occurred in 1972-76 for those aged 60-64, 65-69 and 70-74. The highest rate for those aged 75-79 was in 1977-81, whilst incidence continued to increase until 1989 in the 80-84 and 85 or over age groups.

Females

The truncated age-standardised rates show that lung cancer mortality was fluctuating in both age groups in the early part of the century (Fig. 5.9). However, after 1916-20 rates in both groups showed a substantial and steady rise. In females aged 35-64 mortality rose from 1.81 per 100,000 in 1916-20 to 29.90 (95%CI 29.40 to 30.40) in 1986-90 (Table 5.1). There has been a subsequent decrease to 28.20 (95%CI 27.20 to 29.30) per 100,000 in 1991. In the older age group, mortality rose from 3.31 to 158.00 (95%CI 156.00 to 159.00) per 100,000 between 1916-20 and 1986-90, with a continued increase to 168.00 (95%CI 164.00 to 172.00) in 1991. Incidence rates have also increased in both age groups between 1962-66 and 1982-86 (Figs. 5.10, 5.11), from 18.80 to 33.10 (95%CI 32.60 to 33.60) per 100,000 in younger females and 48.30 to 140.00 (95%CI 138.00 to 141.00) per 100,000 in those aged 65 or over (Table 5.2). A further significant rise to 171.00 (95%CI 167.00 to 175.00) is demonstrated in older females for 1989, whereas an increase to 34.30 (95%CI 33.10 to 35.40) per 100,000 in those aged 35-64 is not statistically significant. Indeed, the increases in this age group since 1977-81 are not significant (Fig. 5.10, Table 5.2).

Table 5.3) reveal different trends for younger and older age groups. In those groups aged 60 or over mortality rates have continued to rise until 1986-90, although there has been a decrease from 109.6 per 100,000 in 1986-90 to 103.3 in 1991 for the 60-64 group. However, mortality has been decreasing since 1961-65 (Table 5.5) in cohorts born subsequent to about 1926 (Fig. 5.12). Similarly, although increases in incidence for the 60 or over age groups have continued up to 1989 (Table 5.3), there is a suggestion of a fall for cohorts bom after about 1927 (Fig. 5.13). However, fluctuations in the rates of the 35-39 age group are to be noted.

5.3.1.2. Geographic distribution

In males, the SIRs for all RHAs were either significantly higher or lower than the national ratio of 100 (Fig. 5.14, Table 5.6). The highest ratios were demonstrated in the Northern and North Western RHAs at 120 (95%CI 118 to 122), and Mersey 116 (95%CI 113 to 119). The lowest ratios of 82 (95%CI 80 TO 83) and 86 (95%CI 84 to 89) were in the South Western and Wessex RHAs respectively. In all other RHAs the ratios were within the range 90 to 109. Similarly, amongst females the highest ratios of 120 (95%CI 116 to 125), 119 (95%CI 115 to 124) and 112 (95%CI 108 to 115) were also demonstrated in the Mersey, Northern and North Western RHAs (Fig 5.15, Table 5.6). Although a low SIR of 82 (95%CI 79 to 85) was found for the South Western RHA, the lowest ratio demonstrated was in the West Midlands RHA at 79 (95%CI 77 to 82). However, a low SIR of 88 (95%CI 85 to 92) was also noted in Wales. The SIRs in all remaining RHAs were within the range 90 to 109, although in the North West Thames, North East Thames, South West Thames and Wessex RHAs

they showed statistically significant differences from the national ratio.

In both genders, lung cancer incidence was greatest in the north and north-western areas of England and low in the south-west (Fig. 5.16). However, there was also low female incidence in Wales and the west midlands. The degree of association between the SIRs of each gender is reflected by a Spearman’s Rank Correlation Coefficient of 0.66 (95%CI 0.22 to 0.87. p<0.01).

5.3.2 Liver cirrhosis

5.3.2.1 Temporal trends

In 1911-15 the liver cirrhosis age-standardised mortality rates for males and females stood at 14.54 and 9.75 per 100,000 respectively (Fig. 5.17, Table 5,7). There was a substantial decrease in mortality for both genders in 1916-20, to 8.32 per 100,000 in males and 4.04 in females. Subsequently, reductions in mortality continued until 1941- 45 in females, falling to 1.02 per 100,000. However, in males the lowest rate of 2.03 per 100,000 occurred five years later in 1946-50. Following an increase to 2.36 in 1951-55, male mortality remained relatively stable until 1971-75 when the rate increased to 2.80 per 100,000. This rise in mortality continued, to reach rates of 4.59 in 1986-90 and 5.18 per 100,000 in 1991. Similarly, after 1941-45 female rates increased to 1.53 per 100,000 in 1951-55, remained relatively stable until a rise to 2.10 in 1971-75 and have since increased to 3.19 per 100,000 in 1986-90 and 3.31 in

In both genders liver cirrhosis mortality initially increases with age (Fig. 5.18, Tables 5.8 & 5.9). However, in females peak mortality occurs in the 65-69 age group, with lower rates in those aged 70-79 and a substantial decrease in the 80 or over group. Similarly, males aged 80 or over have a substantially lower mortality rate than those with peak mortality aged 65-69 to 75-79.

Males

Truncated age-standardised rates reveal decreasing mortality in both age groups during the earlier part of the century (Fig. 5.19, Table 5.7). Mortality fell from 32.60 per

100,000 in 1911-15 to 3.63 (95%CI 3.44 to 3.82) in 1946-50 for those aged 35-64, whilst in the older group the reduction was from 57.80 to 11.10 (95%CI 10.50 to

11.80) per 100,000 in the same period. Increases occurred in both groups in 1951-55; to 4.26 (95%CI 4.06 to 4.46) in younger males and 12.50 (95%CI 11.80 to 13.20) in those aged 65 or over. Subsequently, rates in both age groups remained relatively stable, until 1966-70 in males aged 35-64 and 1971-75 in the older group. In the younger age group, the mortality rate rose to 5.82 (95%CI 5.60 to 6.04) per 100,000 in 1971-75 and has continued to increase to 9.94 (95%CI 9.64 to 10.20) in 1986-90, and 11.50 (95%CI 10.80 to 12.20) per 100,000 in 1991. For males aged 65 or over, mortality increased in 1976-80 to 13.70 (95%CI 13.10 to 14.30) per 100,000 and has risen to 19.20 (95%CI 18.50 to 19.90) in 1986-90 and 20.30 (95%CI 18.80 to 21.90) per 100,000 in 1991.

Age-specific rates confirm decreasing mortality in all age groups during the earlier part of the century (Fig. 5.20, Table 5.8). This continued until 1941-45 for males aged 35-

39 or 40-44, and to 1946-50 for the age groups 45-49 to 60-64. Subsequent to these time periods there were initial increases in mortality, followed by slightly fluctuating or relatively stable rates for these age groups until 1966-70. In 1971-75 increases in mortality occurred for all groups aged 35-39 to 60-64; these have been sustained until

1991. Similar trends are apparent for older age groups, although increases in mortality did not occur until 1976-80 in males aged 65-69 to 75-79 and 1981-85 for those aged 80 or over.

Females

The truncated age-standardised rates demonstrate decreasing mortality in both age groups, until 1941-45 in females aged 35-64 and 1946-50 in the older group (Fig. 5.21, Table 5.7). In younger females, the rate decreased from 23.00 per 100,000 in 1911-15 to 1.86 (95%CI 1.73 to 1.98). An increase in mortality occurred in 1946-50, the rate rising to 2.18 (95%CI 2.05 to 2.32) per 100,000. There has been a continued and sustained increase since, to 6.77 (95%CI 6.53 to 7.01) in 1986-90. In females aged 65 or over the rate fell from 33.00 to 4.61 (95%CI 4.24 to 4.98) per 100,000 in 1946- 50. A substantial increase in mortality occurred in 1951-55, the rate rising to 7.91 (95%CI 7.45 to 8.37); this has also continued until 1986-90 increasing to 13.60 (95%CI 13.10 to 14.10) per 100,000.

Age-specific rates confirmed the decreasing trend in mortality for all age groups in the earlier part of the century (Fig. 5.22, Table 5.9). However, about 1950 the trend was reversed, commencing in 1946-50 for those groups aged up to 60-64 and in 1951-55 for groups aged 65-69 and over. Mortality rates then remained stable or fluctuated

slightly until 1971-75 when more pronounced increases in mortality occurred in age groups up to 60-64; these have continued to 1986-90. In older age groups, this upward trend commenced in 1976-80 for females aged 65-69 and 1981-85 for those aged 70- 74 to 80 or over.

5.3.2.2 Geographic distribution

Significantly raised male SMRs were demonstrated in four RHAs: Northern 127 (95%CI 115 to 141), Wales 120 (95%CI 108 to 134), Mersey 119 (95%CI 105 to 133) and North Western 111 (95%CI 101 to 122) (Fig. 5.23, Table 5.10). In a further four RHAs significantly low ratios could be delineated, the lowest being 70 (95%CI 58 to